OBJECTIVES: Cardiovascular disease is the most common extrahepatic complication and leading cause of death in metabolic dysfunction‑associated fatty liver disease (MAFLD). However, early metabolic biomarkers for subclinical atherosclerosis (SA) in MAFLD patients remain unclear. This study used untargeted metabolomics to investigate serum metabolite and pathway changes associated with SA in MAFLD patients and to screen for potential serum metabolic biomarkers. METHODS: From December 2023 to December 2024, 64 MAFLD patients with SA enrolled in the "Pan-Vascular Cohort Study" at the Health Management Medical Center of the Third Xiangya Hospital of Central South University were included as an experimental group, and 50 age- and sex-matched MAFLD patients without SA were included as controls. MAFLD was diagnosed according to the Guidelines for the Prevention and Treatment of Metabolic (Non-Alcoholic) Fatty Liver Disease (2024 Edition). SA was defined by the presence of multi-segment arterial plaques and/or peripheral arterial obstruction. Untargeted metabolomics profiling was performed using ultra-high performance liquid chromatography coupled with mass spectrometry. Differential metabolites were screened using fold change (FC) analysis, principal component analysis, and orthogonal partial least squares discriminant analysis. Receiver operating characteristic (ROC) curves evaluated diagnostic performance of differential metabolites, and pathway enrichment analysis identified metabolic pathways associated with MAFLD and atherosclerosis. RESULTS: Using selection criteria of FDR-corrected q<0.05 and FC >1.5 or <0.67, a total of 415 differential metabolites were identified between the 2 groups. The top 5 differential metabolites were lysophosphatidylcholine (18꞉3), chelerythrine, N,N-hexamethylene thiocarbamate S-ethyl ester, N-ethyl-o-crotonylmethylaniline, and lysophosphatidylcholine (18꞉1). ROC analysis based on support vector machine (SVM) modeling of these top 5 metabolites yielded an area under the curve of 1. Differential metabolites were mainly enriched in histidine metabolism, sphingolipid metabolism, glycerophospholipid metabolism, unsaturated fatty acid biosynthesis, tryptophan metabolism, and taurine/hypotaurine metabolism. CONCLUSIONS: The metabolic profiles of MAFLD patients with and without SA differ significantly. Key differential metabolites may serve as potential biomarkers for predicting atherosclerosis in MAFLD patients. 目的: 代谢相关脂肪性肝病(metabolic dysfunction‑associated fatty liver disease，MAFLD)最常见的肝外并发症及死因是心血管疾病，MAFLD患者发生亚临床动脉粥样硬化(subclinical atherosclerosis，SA)的早期代谢标志物目前尚不明确。本研究通过非靶向代谢组学探索MAFLD患者发生SA的血清代谢物及代谢途径的变化，筛选潜在的血清代谢标志物。方法: 选取2023年12月至2024年12月在中南大学湘雅三医院健康管理医学中心“泛血管队列研究”纳入的64例MAFLD合并SA患者作为实验组，同期选取50例年龄、性别匹配的MAFLD未合并SA的受检者作为对照组。MAFLD诊断标准参照《代谢相关(非酒精性)脂肪性肝病防治指南(2024年版)》，采用多血管节段动脉粥样硬化斑块定义SA。使用超高效液相色谱和质谱仪进行非靶向代谢组学检测。采用差异倍数(fold change，FC)分析、主成分分析和正交偏最小二乘判别分析方法筛选差异代谢物，受试者操作特征(receiver operating characteristic，ROC)曲线评价差异代谢物的诊断效能，通路富集分析方法筛选与MAFLD及动脉粥样硬化发生相关的通路。结果: 基于差异代谢物筛选标准错误发现率(false discovery rate，FDR)-corrected q<0.05，FC>1.5或<0.67)，2组间共筛选出415个差异代谢物。排名前5的差异代谢物为溶血磷脂酰胆碱(18꞉3)、白屈菜红碱、N,N-六亚甲基硫代氨基甲酸-S-乙酯、N-乙基-邻巴豆酰甲基苯胺、溶血磷脂酰胆碱(18꞉1)。选取这5个差异代谢物使用支持向量机(support vector machine，SVM)进行建模并绘制ROC曲线，曲线下面积为1。差异代谢物主要富集于组氨酸代谢、鞘脂代谢、甘油磷脂代谢、不饱和脂肪酸生物合成、色氨酸代谢、牛磺酸与亚牛磺酸代谢等通路。结论: MAFLD患者是否发生SA的代谢谱存在显著差异，关键差异代谢物可能是预测MAFLD患者是否发生动脉粥样硬化的潜在生物标志物。.